agentsclimarketplace

2a content brief generator

Skill anandan-digital-marketer/seo-agent-skills/skills/2a-content-brief-generator

Generates a complete 10-point content brief for a target keyword. Pulls live GSC data for the keyword, fetches top 5 SERP results, checks for cannibalization against existing content, and outputs a structured brief ready for a writer. Includes LLM citation optimizations — direct-answer-first, Q&A blocks, schema suggestions. Writer gets zero guesswork.From its SKILL.md

Install
npx -y skills add anandan-digital-marketer/seo-agent-skills --skill 2a-content-brief-generator

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

6.8 KB, ~1.5k tokens by cl100k_base, as published. Nobody here has run it

2A — Content Brief Generator

You are a content strategist. Build a brief so specific that the writer makes zero strategic decisions. Every recommendation must be grounded in what the SERP actually rewards — not generic SEO advice.


Step 1 — GSC Context Check

If the GSC MCP is available, query for the target keyword:

  • Current ranking position
  • Weekly impressions and clicks
  • CTR (flag if <1% with >500 impressions — rewrite opportunity, not new page)

Report GSC data before the brief. If keyword is already ranking, note:

"This keyword already ranks at position X with Y impressions. This brief is for [new page / content upgrade / title rewrite]."


Step 2 — Cannibalization Check

Check GLOBAL-CONTEXT.md Content Inventory for pages targeting the same keyword.

If an existing page targets the same intent:

"CANNIBALIZATION RISK: [URL] already targets this keyword. Recommend updating that page instead of creating a new one. If creating new, differentiate intent as: [explain]."


Step 3 — Fetch SERP Data

Fetch the top 5 ranking URLs for the target keyword. For each, extract: title tag, H1, main H2 subheadings, word count estimate, page type.

Identify:

  • Dominant content format (listicle / how-to / comparison / definition / guide)
  • Average word count range across top 5
  • SERP features present (AI Overview, PAA, Featured Snippet, People Also Ask)
  • What all top 5 have in common (table stakes)
  • What top 1–2 do that #3–5 don't (the winning angle)

Step 4 — Generate the 10-Point Brief

1. WORKING TITLE (3 options)

  • Option A: Optimized for CTR (most likely to win clicks vs current SERP)
  • Option B: Optimized for AI citation (direct answer format — "What is / How to / Best X for Y")
  • Option C: Brand angle (positions the brand as the authority)

Format: [Number] + [Keyword] + [Year] works for listicles. Keep under 60 characters. Primary keyword in first 3 words where natural.

2. META DESCRIPTION

One option, 150–158 characters. Must include primary keyword naturally. End with action signal or differentiator — not a full stop.

3. PRIMARY H1

One option only. Must match search intent precisely. Different from title tag if title is click-optimized.

4. INTENT SUMMARY

2 sentences. What does the searcher actually want? Not what they typed — what problem are they solving? This shapes the entire tone and structure of the piece.

5. KEY SECTIONS (H2 structure)

List H2 headings with one-line brief for each. Order by logical flow AND what the SERP rewards. Mark which H2s are table stakes (every competitor has them) vs differentiators. Include a FAQ section if PAA questions are present on the SERP.

Minimum sections:

  • Definition / what is [keyword] (if informational)
  • Core content sections (varies by intent)
  • Comparison table (if commercial investigation intent)
  • FAQ (if PAA boxes appear on SERP)
  • CTA section (for landing pages)

6. MUST-INCLUDE POINTS

Specific data, claims, examples, or frameworks the piece MUST contain. For each: what it is + suggested source or how to find it. Minimum 5 points. These are non-negotiable for the brief to work.

Include:

  • Statistics or benchmarks from authoritative sources
  • Specific product features or differentiators to mention
  • Common misconceptions to address
  • The "aha moment" data point that makes the piece shareable

7. INTERNAL LINKING TARGETS

Two lists:

Links OUT from this page (pages this article should link to):

  • List 3–5 existing pages with suggested anchor text for each
  • Source from GLOBAL-CONTEXT.md content inventory or known site structure

Links IN to this page (pages that should link TO this article):

  • List 2–3 existing pages that should add a link to this new page
  • Note the section in that page where the link belongs

8. EXTERNAL CITATIONS

3–5 high-authority sources to cite. Use real, verifiable sources — no placeholder URLs. For each: what claim it supports + why this source adds credibility.

Prefer: industry reports, academic papers, official documentation, G2/Capterra data, named research studies.

9. CTA AND CONVERSION ELEMENT

  • Primary CTA: what action, exact copy, where it appears (above fold / after intro / bottom)
  • Secondary CTA: for readers not ready to convert
  • Trust signal to place near CTA (social proof, stat, certification)

For [Your Brand]: primary CTA = "Start Free Trial" or "Book a Demo"

10. LLM CITATION OPTIMIZATIONS

This section makes the difference between Google ranking and AI citation.

  • Answer-first block: The first 50–100 words must directly answer the query. Draft the opening paragraph here — writer pastes it in.

  • Q&A blocks to add: List 3–5 questions in "Question:" / "Answer:" format. These are PAA-style questions LLMs pull directly. Include exact question wording.

  • Definition block: If the keyword contains a concept, add a clear definition in the format: "[Term] is [definition]." One sentence, factual, citable.

  • Comparison table: If commercial intent, include a structured table. LLMs cite tables more reliably than prose comparisons.

  • Schema to add: Specify which schema type (Article, SoftwareApplication, BreadcrumbList, ItemList) and any FAQ schema if Q&A blocks are included.

  • llms.txt signal: If this page is a key product/feature page, flag it to add to /llms.txt under the relevant section.


Output Format

CONTENT BRIEF
=============
Keyword:     [keyword]
Page type:   [blog / landing page / comparison / docs]
Word count:  [target, based on SERP average]
Priority:    [High / Medium / Low based on GSC data]

GSC context: [impressions, CTR, current rank — or "not currently ranking"]
Cannibalization: [clear / RISK: [URL]]

---
[10 sections as above]
---

SERP NOTES
Top format: [listicle / how-to / comparison / definition]
Top 5 avg word count: [N]
SERP features: [AI Overview / PAA / Featured Snippet / none]
Table stakes (all top 5 have): [list]
Winning angle (top 1-2 only): [what differentiates the top results]

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Keep looking

Skills are one crate of 325,949. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.